Abstract
Background Mitigating the effect of linear transport infrastructure (LTI) on fauna is a crucial issue in road ecology. Wildlife crossing structures (tunnels or overpasses) are one solution that has been implemented to restore habitat connectivity and reduce wildlife mortality. Evaluating how these crossings function for small wildlife has often been recommended but, mainly due to technical limitations, is not often conducted in practice or only as short-term monitoring (less than 1 year). In this study, we developed and tested an automated device that records the detailed behaviour of animals when using wildlife tunnels. The method is based on marking and detecting individuals with RFID (radio-frequency identification) tags and allows small animals to be tracked. Composed of four antennas (detectors) placed at roughly 2 m intervals, the system was tested in a tunnel in northern France in the summer of 2017. One species of amphibian (the toad Bufo spinosus, n = 13) and two carabids (the ground beetles Carabus coriaceus and Carabus nemoralis, respectively, n = 3 and n = 4), marked beforehand with PIT (passive integrated transponder) tags, were recorded when they crossed these detectors. This allowed individual trajectories, including crossing speed, to be estimated. Results We found that 12 of the 13 toads and 3 of the 7 ground beetles successfully crossed the entire wildlife passage of 7 m long. The detection rate of each detector varied from 8.33 to 100%, with a mean of 52.08%. All individuals were recorded by at least one detector. We observed high variability in the crossing characteristics of toads (mean transit duration = 41 min and 15 s ± 25 min) and ground beetles (6 h 11 min ± 3 h 30 min). The system provided information on precise trajectories (e.g., crossing speed, U-turns, distance travelled in the tunnel, proportion of individuals reaching the exit, etc.) for each individual, in a context of tunnel crossing. Conclusion The system allowed us to record small animal behaviour in the context of tunnel crossing in which other types of tracking (e.g. radio-tracking) or detection (e.g. camera traps) are not effective. It also opens the possibility for a range of experiments that would contribute to a better understanding of the behaviour of small animals in tunnels, allowing a comparison of tunnel characteristics (such as size, building material, substrate, etc.) with the aim of increasing wildlife use of the tunnel and proposing guidelines for the construction and maintenance of these mitigation measures.